Random crashing - HW or SW?
 

Hmmm. Been having my pooter crashing randomly for a few months now and it's finally got to the point where I needed to check it out. I'd always had the thought that it was a memory SIM going bad but taking each out in turn shows that either chip can be present and a crash happens.

Oh, HW is an ASUS M/B with AMD XP2500+ (overclocked to 3200+) with 2x Corsair 512MB XMS3200, ATI 9600 XT graphics card and little else! SW is XP Pro with SP3 and latest (last week) patches for XP and office.

No real pointer to 'it started when 'something' was installed/uninstalled'. The blue screen that comes up points to different places where the crash happened (why I suspected memory). Anyone got anywhere specific I could look? I was thinking it *could* be both SIMs gone at the same time (the PC is over 4 years old) but other than that I was just going to pull things out clean them and put them back. The fans are still going when a crash happens etc etc.

Help me anyone ....

Ta

Dave

on my office HP, it would slow down then lock up and beep at me, turned out to be the motherboard.

If its 4 years old what about M/B battery playing up?
Memtest86+ - Advanced Memory Diagnostic Tool you can try this for testing the memory

is your power supply up to the job?

did you check the event logs for system and application?

Logs: nothing repeatable. As I said, sort of 'random' ...

Power supply: Tagan TG430-U15. Less than a year old.

Dave

can you boot into safe mode and leave it to see if it crashes? (device driver issues)

Turn down or stop overclocking, see if that helps.

Would help to know what the BSDs are showing. Can help point ot the problem. Just wondering if something isn't overheating due to the overclocking. Have you cleaned out the machine as of late, lots of dust can clog up fans etc...

Ta for advice. It's been *overclocked* since day 1 and the crashing has only just started. So 4 years of being overclocked means it *probably* isn't that. The fans are all working as well. No undue temperature spikes about the time of the crash (as reported by Asusprobe) BUT since I've played about with the memory/ tested with memtest it seems fine. However, I've also left the side panel off to make life easier so I'm wondering if something could be overheating. Maybe the CPU and NB heatsinks need removing and reinstalling with new paste??

Don't really want to put a new PC together until Nehalem is out ...

Dave

jus run memtest on it, more likely to be the dimms.

Yeah cos running chips above spec doesn't degrade them faster than normal :)

If you have onboard sound/ethernet etc try disabling each one.
Or take the sound card out.
Diable/disconnect CD/DVD drives one at a time.
Check the internal connections, I've seen a loose sata connection cause random lock ups.
Does it happen at the point a screen saver would start?
When doing large file copies?
When 3d applications start?

If there is no real pattern I'd bet on electrical, have you got a UPS or something to protect against spikes? (I've seen "new" fridges send massive spikes though ;) )

Cheers

Dan

No real pattern BUT I put the side panel back on and hey presto it started crashing again ....

Took it off and vacuumed out the insides - it wasn't too bad anyway as I'd done it just a few weeks ago. Also re-routed a couple of disk cables (power and data) so that they has less chance of getting 'pushed' by the side panel (*maybe* any vibration was transmitted through the side panel to cable to .... it's a theory anyway!).

Also removed memory and graphics card and hoovered them clean + used a pencil eraser to clean the contacts. I also found a whole pile of fluff in a PCI slot that was directly under the graphics card fan. Hmmmm. I wonder .... I also moved a firewire connector from the slot almost next to the graphics card to be as far away as possible.

So far no crashes......

Dave